Our team of climate scientists are taking the stage at the European Geosciences Union (EGU) General Assembly 2022 in Vienna, Austria, from 23 - 27 May 2022.
The EGU General Assembly 2022 brings together geoscientists worldwide for one meeting covering all disciplines of the Earth, planetary, and space sciences. The EGU provides a forum where scientists can present their work and discuss their ideas with experts in all fields of geoscience.
Quick overview of the sessions we're holding:
An interdisciplinary approach to projecting multiple climate-related risks and impacts
Dr Claire Burke - Monday, 23rd May 2022
With ever-increasing risks and impacts from climate change, there is an urgent need for adaptation information which is relevant and useful to policymakers, businesses and the general public.
At Climate X we use an interdisciplinary, impacts-motivated approach to adaptation; combining multiple climate and hazard models to give a holistic view of risk, and engaging end-users at every stage.
Our first version product can project the risks and impacts of climate change-related pluvial and fluvial flooding, extreme heat, landslides, subsidence, and sea-level rise, all at street level UK-wide. We quantify these risks and the financial costs they could incur under low (RCP 2.6) and high (RCP 8.5) emissions scenarios out to 2080.
We deliver risk and impact assessments via an easy-to-use interface, along with relevant and decision-able risk summaries. Aligning robust science at scale with user requirements and expectations is not without its challenges. I will outline our approach to multi-hazard climate risk modelling, and discuss some of the successes and challenges we have had in developing a tool which is aligned with the needs of stakeholders, businesses and other end users.
Regional landslide susceptibility mapping using tree-based machine learning techniques
Dr Hamish Mitchell - Tuesday, 24th May 2022
The identification of assets susceptible to landslide-related damage is critical for planners, managers, and decision-makers in developing effective mitigation strategies. Recent applications of machine learning and data mining methods have demonstrated their use in geotechnical assessments including the spatial evaluation of landslide susceptibility.
At Climate X, we utilise tree-based machine learning techniques alongside geographic information system and remote sensing data to map landslide susceptibility across Great Britain. We compile several conditioning factors—including topographic, subsurface, land use, and climate-related data—and combine them with over 18,000 landslide instances, recorded in National Landslide Database. We evaluate the capabilities of several techniques including, decision tree, bagged tree, random forest, and balanced random forest (applies random undersampling of the majority, non-landslide class) for landslide susceptibility modelling.
Several performance evaluation indices (area under receiver operator characteristic curve (AUC), precision, recall, F1 score) were used to assess and compare the performance of models. We show that the random forest is the most accurate of our models with an AUC of 94.7%.
Our results demonstrate that tree-based algorithms form a robust method to analyse regional landslide susceptibility and provide new insights into locations susceptible to landslide-related damage across Great Britain.
Establishing the potential impacts of climate change on extreme flood events across the UK
Laura Ramsamy - Wednesday, 25th May 2022
The severity and frequency of extreme flood events has intensified both globally, and across the UK. Climate change will influence weather patterns across the UK, making it increasingly important to understand the impacts this may have on future flood events.
We developed 90m hydraulic models to simulate extreme pluvial and fluvial flood events across the UK based on observed events. The models have been climate conditioned, allowing the potential impacts of climate change on extreme pluvial and fluvial flood events to be understood. Using different climate scenarios, we examine the variation in outcome depending on what efforts are taken to reduce emissions.
Modelling the impacts climate change could have on flooding at a national scale, enables us to understand the spatial-temporal distribution of flood risk. This information can be used in the real world for decision making and providing a way to mitigate against the impacts of climate change.
Using remote sensing and GIS to project climate risk for asset management users
Dr James Brennan - Wednesday, 25th May 2022
At Climate X we are producing risk estimates for the UK to help businesses and communities mitigate and adapt for climate change-related losses.
Climate X provides risk scores and expected financial losses from a plethora of hazards including flooding, subsidence, landslides, drought, fire and extreme heat. To do this at the scales we need, Earth Observation (EO) and other geospatial data sets play a crucial role in both physical modelling and risk estimation.
Generating rich geospatial datasets to sit as the bedrock of risk models requires intelligent use of multiple data sources, involving the fusion of EO data from synthetic aperture radar, lidar and optical instruments and across processing levels from L1 to L3. This talk will cover the generation and use of these datasets that drive physical risk models (flooding) as well as ML-enabled models (Landslides and subsidence).
Meeting the demand for multi-hazard climate risk information tailored to financial services
Markela Zeneli - Thursday, 26th May 2022
As uncertainty around the impacts of climate change become more apparent, businesses and communities are relying on cutting-edge information to help them navigate their next steps. Climate X are a climate risk information provider that aims to help businesses and communities prepare for a rapidly changing environment, with an explainable and transparent method.
Our flagship product, Spectra, presents users with a multitude of potential hazards including flooding (fluvial, pluvial, and coastal), subsidence, landslides, and extreme heat. Each hazard risk is quantified at street level, and we project risks and impacts for low emissions (RCP2.6) and high emissions (RCP8.5) scenarios. This allows users to see the difference between the best-case and worst-case scenarios for assets across the UK.
This poster will cover our methods of finding data, interpolating, modelling, and predicting, as well as a tour of our easy-to-use UI.
Projecting losses due to extreme weather events linked to climate change
Kamil Kluza - Thursday, 26th May 2022
Climate change brings unprecedented risks to the future stability of global financial systems and our society. Central Banks and Governments around the world have joined forces to shape new policies in response to risks driven by climate change. These regulations will require firms to proactively identify, model, quantify and manage climate-related risks for the first time.
Climate X have just completed their first of a kind Integrated Assessment Model evaluating asset-level impacts of climate-related hazards across all 22 million addressed buildings in the UK. Each building has a modelled probability, severity and a simple A-F climate rating as well as a projected loss under a given scenario.
Our geospatial core comprises of UK-specific physical risk models including flooding (pluvial, coastal, fluvial) and geohazards (subsidence and landslides). The models are at 90mx90m resolution feed UKCP18 climate scenarios of RCP8.5 and RCP2.6. Flood models are physics-based and reach a Critical Success Index (CSI) of 75%+. Geohazard models use a combination of DinSAR and machine learning modelling with 90%+ accuracy levels (measured by AUC).
Loss models combine the geospatial hazards with buildings’ exposure (square meterage) and respective vulnerabilities: age, use, material built etc. They then apply insurance-based damage curves to compute structure & content losses against building replacement costs.